Influence of Monomer Concentration and Dielectric Constant on

710

Macromolecules 2000, 33, 710-716

Influence of Monomer Concentration and Dielectric Constant on Transfer Reactions in the Carbocationic Polymerization of Indene M. Givehchi, M. Tardi, A. Polton, and P. Sigwalt* Laboratoire de Chimie Macromoleculaire, U.M.R. 76104, Place Jussieu, 75005 Paris Cedex 252, France Received August 4, 1999; Revised Manuscript Received November 24, 1999

ABSTRACT: The cationic polymerization of indene initiated with cumyl chloride and stannic chloride in solution in methylene chloride at -15 °C is controlled up to Mn ) 20 000, with a low transfer to monomer constant (ktrM/kp ) 9 × 10-4). Experiments carried out in compensated media allowed kinetic measurements and showed that the sharp decrease of the Mn at high indene concentrations results from zero order transfer due to the change of dielectric constant (ktr/kp ) 1.5 × 10-3 for a dielectric constant of 5.4). In methylene chloride, for indene concentrations higher than 2 mol‚L-1, the heat release of the fast reaction adds up to the zero order transfer to lower the molar masses. Bimodal distributions result from propagation on free ions and ion pairs.

Introduction The control of cationic polymerizations has been achieved in the last 20 years for a variety of systems, the pioneering work being done by Higashimura1 and Kennedy.2 Several reviews have been published on this topic,3-5 and the possibilities and limitations of these systems have been discussed.6-8 The control generally realized is that of the Mn which corresponds to the theoretical DPn (DPn ) ∆ [M]0/[I]0) and grows linearly with monomer conversion up to values which may vary with the initiating system and with temperature (e. g. Mn linear (Mn lin) up to 2 × 104). Most experiments were made with all the monomer added at the beginning (all monomer in, AMI), others with incremental monomer additions (IMA). In this last case, this shows the near absence of chain-breaking reactions (transfer and termination) in this range of Mn, and during the synthesis time, the system may be considered to be living. Control of the Mn results from quantitative and preferably fast reaction of an initiator (e.g., an alkyl halide) with a Lewis acid activator and from absent or negligible irreversible termination. Termination is reversible and involves the establishment of an equilibrium between active and dormant end groups. In these reactions, transfer is not suppressed, which limits the range of molar masses for which the process is controlled.8,9 Among the various Lewis acids used as activators, titanium tetrachloride is one of the most frequently employed, particularly for the polymerization of isobutylene and arylalkenyl monomers. In the case of the polymerization of indene, Kennedy observed partial initiation of the polymerization with this Lewis acid and 2-chloro-2,4,4-trimethylpentane10 (Mn > Mn theor). Controlled polymerization of indene was obtained for Mn up to 100 000 with methyl cumyl ether as initiator at low temperature.11 With cumyl chloride (CumCl), satisfactory control was obtained at -40 °C in the presence of additives.12 However, the polymerizations were very fast, completed in a few seconds even in the presence of very low concentrations of initiating species, which did not allow us to monitor the evolution of the molar masses during the reaction.13 Besides, the slow exchange between active and dormant forms was the

cause of the rather broad molecular weight distributions (MWD) observed (Mw/Mn ) 2). Consequently, it seemed interesting to check whether other initiating systems would yield lower polymerization rates which would permit to follow the variation of the Mn with conversion and eventually to lower the Mw/ Mn. Stannic chloride, which has a long history as a an initiator/co-initiator for a large number of cationic polymerizations, has recently been used for the controlled polymerization of styrene. Controlled polymerization of styrene (with Mn lin up to 10 000) has been achieved at -15 °C in methylene chloride solution, with phenylethyl chloride as initiator.14 Addition of tetra-nbutylammonium chloride was necessary to avoid bimodal distributions, which were attributed either to the participation of free ions or to competitive initiation by residual water, which was assumed to yield high molar masses. The role of the salt has been investigated.15 Less satisfactory results were obtained with methanesulfonic acid as an initiator,16 but control was achieved (Mn lin up to Mn ) 10.000), in the absence of salt, by operating in a solvent of lower dielectric constant (CHCl3; ) 4.8).17,18 The molecular weight distribution was bimodal in pure CH2Cl2 and became narrower as the proportion of CH2Cl2 in a CH2Cl2/ CHCl3 mixture was decreased. The same observation was made with 4-methylstyrene.19 It should be noted that in all cases, the range of molar masses was lower than 10 000 ([M]0/[C]0 ) 100) and that, under these conditions, a transfer constant to the monomer equal to 10-3 would yield a value of the molar mass less than 10% lower than the theoretical value. On the other hand, the polymerization rates were rather low, and it could be expected that indene would also yield controlled polymerizations with lower reaction rates with the CumCl/SnCl4 initiating system. Experimental Section The reagents were purified under vacuum on sodium films according to procedures previously reported,20 and stored in a vacuum line fitted with “Rotaflo” stopcocks. 2,6-Di-tert-butyl4-methylpyridine (DBMP) was sublimed under vacuum before use. The polymerizations were carried out in sealed tubes and terminated by introduction of cooled mixtures of methanol and methylene dichloride. A few experiments were made in a

10.1021/ma9913010 CCC: $19.00 © 2000 American Chemical Society Published on Web 01/20/2000

Macromolecules, Vol. 33, No. 3, 2000

Carbocationic Polymerization of Indene 711

Table 1. Initiation of the Polymerization of Indene by Various Initiators at -15 °Ca' initiator

T (°C)

Mn

Mw/Mn

CumOMe CumOMe CumOMe CumCl CumCl CumCl 1-chloro-1-phenylethane

-70 -40 -15 -70 -40 -15 -15

45 400 26 400 20 900 21 000 20 500 20 800 23 400

1.90 2.39 2.41 1.70 1.65 1.54 2.18

a [indene] ) 0.43 M; [initiator] ) 2.5 × 10-3 mol‚L-1; [SnCl ] ) 4 2 × 10-2 mol‚L-1. Reaction time: 6 min (complete conversion; calculated Mn ) 20 000).

drybox and yielded similar results. The polyindenes were recovered by precipitation in methanol and dried overnight under vacuum. The conversions were determined from the weight of polymer. The molar masses were measured by size exclusion chromatography, with a set of 105, 104, 103, 500, and 100 Å Waters microstyragel columns using a polystyrene calibration. Comparative measurements by osmotic pressure and matrix assisted laser diffusion spectrometry (in the low molar mass range) confirmed the validity of the values derived from SEC.

Results and Discussion Quantitative initiation is the first prerequisite for the control of a polymerization. Cumyl derivatives have been used as initiators for a large number of systems, in particular cumyl chloride and cumylmethyl ether which proved efficient in the case of the controlled polymerization of indene with TiCl4 and TiCl3OBu.11-13 To check the efficiency of these compounds, a series of polymerizations have been carried out with these two initiators (2.5 × 10-3 M) in the presence of SnCl4 (2 × 10-2 M) at variable temperature. In all cases, the reactions were fast and complete conversion was reached after 30 s. CumOMe is not an efficient initiator with SnCl4 at -70 °C; the molar mass of the polymer is about twice the calculated value for a controlled process (Table 1). The molar mass decreases when the temperature is increased and corresponds to the expected value at -15 °C, which is in agreement with a quantitative initiation at this temperature. With cumyl chloride, the molar masses have the expected value over the whole range of temperature, which shows that initiator efficiency is 100% even at -70 °C, contrary to the case of ionization by TiCl4, and that initiation is fast enough compared to propagation. This initiator has been used in the following experiments. 1-phenylethyl chloride, which has been used as an initiator for the polymerization of styrene, is also efficient at -15 °C for indene, but initiation is apparently slower (Mn exptl are slightly higher than Mn theor). (1) Polymerizations in Dichloromethane Solution. To check the range of [M]0/[I] o ratios for which control is effective, a series of polymerizations has been carried out at variable initial indene concentrations ([SnCl4 ] ) 2 × 10-2 M; [CumCl] ) 2.5 × 10-3 M) at -15 °C. Polymerizations were stopped after 6 min, but complete conversion was reached already after 30 s. Consequently, the only mass measurements which could be done were on polyindenes resulting from polymerizations at complete conversion. The results are reported in Figure 1 and Table 2. A linear increase of the molar masses, corresponding to the expected values, is observed for indene concentrations lower than 0.5 M ([M]0/ [I]0 ) 200). In the case of the polymerization of styrene, the domain of control is restricted to monomer-to-

Figure 1. Polymerization of indene initiated with CumCl/ SnCl4. Evolution of the molar masses with the initial indene concentration. Reaction time: 6 min, complete conversion. [CumCl] ) 2.5 × 10-3 mol‚L-1; [SnCl4] ) 2 × 10-2 mol‚L-1; solvent, CH2Cl2. T: -15 °C. Key (-) controlled process; (- -) calculated values for CM ) 9 × 10-4; (∆) experimental points. Table 2. Polymerization of Indene Initiated with Cumyl Chloride at -15 °C, Showing the Influence of the Monomer Concentration on the Molar Massesa expt no.

[indene]o (mol‚L-1)

Mn exptl

Mn calcd

Mw/Mn

1 2 3 4 5 6 7 8

0.215 0.43 0.43 0.86 1.25 1.72 2.0 2.5

10 500 20 800 21 000 31 900 33 000 27 900 14 500 12 100

10 000 19 900 19 900 39 900 58 100 79 900 92 900 116 100

1.52 1.54 1.74 1.74 2.24 2.02 3.46 2.96

a [CumCl] ) 2.5 × 10-3 mol‚L-1; [SnCl ] ) 2 × 10-2 mol‚L-1; 4 solvent ) CH2Cl2. Reaction time: 6 min (complete conversion).

Figure 2. Polymerization of indene initiated with CumCl/ SnCl4. [CumCl] ) 2.5 × 10-3 mol‚L-1; [SnCl4] ) 2 × 10-2 mol‚L-1; solvent, CH2Cl2. T: -15 °C. Transfer constant to indene.

initiator ratios lower than 100.17 In the present case, control is still effective for a molar ratio of about 200. In this range of concentrations, a plot of 1/DPn vs 1/DPo (DPn and DPo: experimental and theoretical degrees of polymerization) yields a straight line with an intercept of 9 × 10-4 (Figure 2), which is the value of the transfer constant (CM ) ktrM/kp) to indene under these conditions. Between 0.43 and 0.86 M, the experimental masses are in agreement with those calculated taking this

712

Givehchi et al.


Table 3. Polymerization of Indene at a Fixed Dielectric Constant (Indene + Methylcyclohexane, 30% v/v; CH2Cl2, 70% v/v; E ) 7.2), Showing the Evolution of the Molar Mass with the Initial Monomer Concentrationa expt no.

[indene] (mol‚L-1)

Mn exptl

Mn calcd

Mw/Mn

1 2 3 4 5

0.21 0.43 0.85 1.50 2.20

9 400 10 000 14 000 16 300 19 100

9 750 20 000 39 500 69 700 102 200

2.23 2.24 2.11 2.16 2.47

a [SnCl ] ) 2.10 × 10-2 mol‚L-1; [CumCl] ) 2.5 × 10-3 mol‚L-1; 4 T ) -15 °C. Reaction time: 6 min (complete conversion).

transfer into account (Figure 1). However, for higher indene concentrations a decrease from Mn ) 33 000 for 1.25 M indene down to Mn ) 12 000 for 2.5 M indene is observed which cannot be accounted for by classic transfer to monomer alone, since if this transfer were preponderant, the molar masses should remain practically constant. To investigate the existence of transfer, a polymerization has been carried out at high indene concentration under the same conditions as in experiment 8 of Table 2 ([indene] ) 2.5 M; [CumCl] ) 2.5 × 10-3 M; [SnCl4] ) 2 × 10-2 M), but in the presence of 2,6-di-tert-butyl4-methylpyridine (4 × 10-3 M). The reaction stops at 11% yield; the molar mass of the polyindene (Mn ) 11 300; Ip ) 2.05) corresponds to a concentration of macromolecules of 2.6 × 10-3 M, equivalent within experimental error to the concentration of the initiator. This molar mass corresponds to that of experiment 8 of Table 2 (without hindered pyridine) which went to completion (Mn exptl ) 12 100). This suggests that, in the former case, the polymerization had stopped, either by capture by the hindered pyridine of a proton resulting from transfer or by termination of the active centers by the pyridine. The similar values of the Mn in the two experiments suggests that the former explanation is the most probable. Assuming the transfer constant to monomer determined above (9 × 10-4), the molar mass for [indene]0 ) 2.5 M should have been 61 000 at complete conversion. This implies that some other type of transfer takes place at high monomer concentration. (2) Polymerizations in Media of Fixed Dielectric Constant. The main change at high indene concentration is the decrease of the dielectric constant of the medium as the proportion of indene ( ) 2) with respect to CH2Cl2 ( ) 9) increases. To check whether a similar evolution of the molar masses would be observed in media of constant polarity, experiments have been carried out at constant volumes of CH2Cl2 and of indene + methylcyclohexane (MeCH). In the case of the polymerization of styrene, a medium of lower dielectric constant (CHCl3 instead of CH2Cl2) allowed control of the molar masses up to Mn ) 10 000 without addition of salt.19 For a dielectric constant equal to 7.2 at 20 °C (30% v/v of indene + methylcyclohexane in CH2Cl2) the reaction is still rapid and the molar masses reported are from reactions carried out to completion. The molar masses increase with the initial indene concentration (Table 3), but are much lower than the calculated values for a controlled process (Figure 3) and than those obtained in pure CH2Cl2. However, there is no decrease of the Mn at high indene concentration (2.2 M), as was the case previously. In a medium of lower dielectric constant (50% v/v of indene + methylcyclohexane in CH2Cl2; ) 5.4), the

Figure 3. Polymerization of indene initiated with CumCl/ SnCl4 in compensated media. [CumCl] ) 2.5 × 10-3 mol‚L-1; [SnCl4] ) 2 × 10-2 mol‚L-1. T: -15 °C. Key: (∆) ) 7.2 (indene + MeCH, 30% v/v in CH2Cl2); variable initial indene concentration; reaction time, 6 min; complete conversion. (O) ) 5.4 (indene + MeCH: 50% v/v in CH2Cl2); [indene] ) 0.43 M; variable conversion. (+) ) 5.4 (indene + MeCH: 50% v/v in CH2Cl2); variable initial indene concentration; reaction time, 6 min; conversion, 36%; (-) calculated Mn for a controlled process. Table 4. Polymerization of Indene at a Fixed Dielectric Constant (Indene + Methylcyclohexane, 50% v/v; CH2Cl2, 50% v/v; E ) 5.4), Showing the Evolution of the Molar Mass with Conversiona reacn expt time global no. (min) convn Mn 1 2 3 4 5

2 10 15 30 55

0.15 0.42 0.52 0.72 0.89

2500 5000 5600 6400 6900

high Mn

low Mn

39 700 2100 20 500 5000 19 900 5900 6400 6900

Mw/Mn Mw/Mn high Mn low Mn 1.55 2.02 1.96

1.24 1.49 1.56 2.20 2.28

calcd Mn 3000 8400 10400 14400 18000

a [indene] ) 0.43 mol‚L-1; [SnCl ] ) 2.10-2 mol‚L-1; [CumCl] 0 4 ) 2.5 10 -3 mol‚L-1; T ) -15 °C.

polymerization is slow enough to allow monitoring the evolution of the molar masses with conversion. The results of a polymerization ([indene] ) 0.43 M; [CumCl] ) 2.5 × 10-3 M; [SnCl4] ) 2 × 10-2 M) are reported in Table 4. The consumption of indene follows a first-order law (Figure 4). However, up to 52% yield, the molecular weight distribution (MWD) is bimodal, with a small peak of high Mn and a main peak of low Mn (Figure 5). The ratio of the two peaks, which can be measured up to 52% yield by integration of the SEC traces, remains constant (18% of high Mn polymer). For higher conversions, the bimodality is no longer visible, but the two peaks merge, as is shown by the broadening of the MWD (Table 4; Figure 5). The molar masses are much lower than those obtained in pure CH2Cl2 and in a medium of dielectric constant ) 7.2. A polymerization carried out under the same conditions ([indene]0 ) 0.43 M), but in the presence of 1.6 × 10-3 M 2,6-ditertbutyl-4-methylpyridine (yield, 9% in 2 min; Mn ) 2700; Mn theor ) 1800) yielded a polymer having a unimodal distribution with a narrow polydispersity (Mw/Mn ) 1.22) with no visible trace of high molar mass polymer (Figure 5). In the same medium ( ) 5.4), for polymerizations at variable initial indene concentration, the yield in 6 min was 36% in all cases independently of [M]0, which is in agreement with a first-order internal law for the



Figure 4. Polymerization of indene initiated with CumCl/ SnCl4 in a compensated medium ( ) 5.4; indene + MeCH, 50% v/v in CH2Cl2). [indene] ) 0.43 mol‚L-1; [CumCl] ) 2.5 × 10-3 mol‚L-1; [SnCl4] ) 2 × 10-2 mol‚L-1. T: -15 °C. Firstorder plot for the comsumption of indene.

Figure 6. Polymerization of indene initiated with CumCl/ SnCl4 in a compensated medium ( ) 5.4; indene + MeCH, 50% v/v in CH2Cl2). [CumCl] ) 2.5 × 10-3 mol‚-1; [SnCl4] ) 2 × 10-2 mol‚L-1. T: -15 °C. Reaction time: 6 min. Conversion: 35%. Evolution of the SEC traces with the initial indene concentration. [indene]o: (A) 0.21 mol‚L-1; (B) 0.43 mol‚L-1; (C) 1.5 mol‚L-1. Table 6. Incremental Monomer Addition in the Polymerization of Indene in CH2Cl2 Solution, Showing the Influence of the Temperature on the Stability of the Active Centersa

temp (°C)

1st addition (mol‚L-1)

2nd addition (mol‚L-1)

Mn exptl

Mn theor

Mw/Mn

-15 -15 -15 -75 -75

0.215 0.215 0.430 0.215 0.215

0 0.215 0 0 0.215

10 500 14 300 20 800 10 500 21 250

10 000 20 000 20 000 10 000 20 000

1.5 1.9 1.5 2.5 1.5

a

Figure 5. Polymerization of indene initiated with CumCl/ SnCl4 in a compensated medium ( ) 5.4; indene + MeCH, 50% v/v in CH2Cl2). [indene] ) 0.43 mol‚L-1; [CumCl] ) 2.5 × 10-3 mol‚L-1; [SnCl4] ) 2 × 10-2 mol‚L-1. T: -15 °C. Evolution of the SEC traces with conversion. Conversion: (A) 15%; (B) 42%; (C) 89%; (D) [2,6-di-tert- butyl-4-methylpyridine] ) 1.6 × 10-3 mol‚L-1, conversion 9%. Table 5. Polymerization of Indene at a Fixed Dielectric Constant (Indene + methylcyclohexane, 50% v/v; CH2Cl2, 50% v/v; E ) 5.4), Showing the Evolution of the Molar Mass with the Initial Monomer Concentrationa expt [indene]o global no. (mol‚L-1) convn Mn 1 2 3 4 5 6

0.11 0.21 0.43 0.85 1.49 2.00

0.38 0.36 0.35 0.33 0.34 0.40

Mn high mass

Mn Mw/Mn Mw/Mn low high low calcd mass mass mass Mn

2700 16 300 1900 3200 19 000 3300 4000 29 100 3700 5600 5600 6200 6200 6600 6600

1.67 1.58 1.65

1.16 1.21 1.46 2.43 2.26 2.21

1900 3600 7000 12600 23100 37200

a [SnCl ] ) 2 × 10-2 mol‚L-1; [CumCl] ) 2.5 × 10-3 mol‚L-1; T 4 ) -15 °C. Reaction time: 6 min.

consumption of indene, (Table 5). For initial indene concentrations lower than 0.85 M, a bimodal MWD is observed (Figure 6), with a small peak of high molar

[SnCl4] ) 2 × 10-2 mol‚L-1; [CumCl] ) 2.5 × 10-3 mol‚L-1.

mass and a main peak of low Mn. For higher concentrations, the two peaks merge, as is shown by the broadening of the MWD. The Mn of the low molar mass peak increases with the amount of monomer consumed, but are lower than the calculated values (Figure 3). The Mn of the high molar masses peak also increase (Table 5). Successful IMA is the best proof for livingness. IMA have been carried out at -15 and -75 °C (Table 6). The second monomer addition was made 6 min after the first polymerization, long after the polymerization time (30 s). At -75 °C the Mn of the final polymer had the expected value for a living process, the polydispersity decreases after the second addition, and the dormant species could be reactivated. At -15 °C, the Mn is lower (14 300 instead of 20 000). This shows that at least a fraction of the end groups has been destroyed after 6 min. This destruction probably takes place after the polymerization through ionization of chlorinated chains by SnCl4, followed by proton abstraction from the β-carbon in the absence of monomer. This would ultimately yield unsaturated dead chains, HCl and SnCl4, which would initiate new chains upon a second addition of monomer. However, the polyindene of an AMI (all monomer in) addition carried out at -15 °C with twice the concentration of the IMA had the expected value. At -15 °C, addition at a shorter time, immediately after the first

714

Givehchi et al.

Figure 7. Polymerization of indene initiated with CumCl/ SnCl4 in compensated media. [CumCl] ) 2.5 × 10-3 mol‚L-1; [SnCl4] ) 2 × 10-2 mol‚L-1. T: -15 °C. Apparent transfer constants to monomer: (∆) ) 7.2 (indene + MeCH, 30% v/v in CH2Cl2); variable initial indene concentration; reaction time, 6 min; complete conversion. (O) ) 5.4 (indene + MeCH, 50% v/v in CH2Cl2); [indene] ) 0.43 mol‚L-1; variable conversion.

reaction, might have yielded the expected value. Anyway, under these conditions, the polymerization is controlled, but not living. (3) Discussion. Polyindenes with controlled molar masses up to Mn ) 20 000 ([M]0/[I]0 ) 200; Mw/Mn ) from 1.57 to 1.74) have been obtained with the SnCl4/ CumCl initiating system at -15 °C, which is better than that for styrene at the same temperature.17 This is probably due to the absence of indanyl-type termination, which imparts a better stability to the cation. Slightly better control of Mw/Mn than with indene and TiCl4 may result from a lower polymerization rate, from a smaller ionization of the end groups, and from more frequent exchanges with the dormant species. Up to [indene]0 ) 0.8 M ([M]0/[I]0 ) 350), the evolution of the Mn with indene concentration is in agreement with a transfer constant to indene equal to 9 × 10-4. Control is lost at higher [M]0/[I]0 ratios and the Mn actually decreases in the range of high indene concentrations. The domain of high monomer concentrations has rarely been investigated in controlled polymerization mainly because control is lost on account of transfer reactions, but the observation of a decrease of Mn at high indene concentrations was made by Sigwalt21 and Mare´chal22 for noncontrolled polymerizations. However, the polymerizations are fast, and complete conversion is obtained in less than 30 s. The reactions are still relatively fast in a compensated medium of slightly lower polarity ( ) 7.2). Under these conditions, the Mn values increase with the [M]0/[I]0 ratio over the whole range of indene concentrations (up to 2.2 M) but are much lower than those obtained in pure CH2Cl2. A lower polarity ( ) 5.4) makes kinetics measurements possible. There is still an increase of the Mn values, both with conversion and with initial indene concentration at a given conversion, but they tend to level off. Transfer increases at low dielectric constant. In all cases, the plots of 1/DPn vs 1/DPo yield straight lines which might be at first sight in agreement with a transfer to monomer. The apparent value of CM increases when the dielectric constant of the medium decreases from ) 9 (CM ) 9 × 10-4 in pure CH2Cl2, Figure 2) to ) 7.2 (CM ) 5 × 10-3) and to ) 3.5 (CM ) 1 × 10-2, Figure 7).


The main difference between the polymerization in pure CH2Cl2 and in the compensated media is the evolution of the molar mass which decreases with increasing concentration of indene in CH2Cl2 above 1 mol‚L-1 while they still increase up to [M]0 ) 2.2 mol‚L-1 even for a relatively high polarity ( ) 7.2) and tend to level off for ) 5.4. In all cases, lowering the dielectric constant decreases the molar masses, indicating an increase of the transfer reactions. The influence of the dielectric constant on the various reactions which may take place during a polymerization has been discussed.23 A decrease of the dielectric constant has a negative effect on the ionization rate (dipole-dipole reaction leading to ionic species). It increases the rate of ion-ion reactions such as deactivation of the ion pairs and transfer to the counterion, and has little influence on ion-dipole reactions such as propagation and transfer to the monomer. Slower activation and faster deactivation account for the drop of overall polymerization rate at low dielectric constant by shifting the equilibrium toward the dormant species. Although the 1/DPn vs 1/DPo plots for transfer to monomer yield straight lines, this transfer should not be as strongly affected as it is here by a change of dielectric constant. However, transfer to the counterion has to be considered under these conditions. Incidence of zero order transfer with respect to the monomer has been shown in the case of the polymerization of R-methylstyrene in a mixed solvent (CH2Cl2-methylcyclohexane 50/50 v/v)24 and in the polymerization of isobutylvinyl ether in toluene.25 Recently, transfers to the counterion have been measured in the case of the polymerization of isobutylene in mixed solvents.26 Transfer decreases at low temperatures and at high dielectric constant. Assuming the existence of transfer to the counterion and to the monomer (which is still operative), the DPn can be calculated according to the following relationship:

DPn )

[M]OY [I]o + CM[M]OY + Ctr ln[1/(1 - Y)]

Hence:

[M]OY - CM[M]OY ) [I]o + Ctr ln[1/(1 - Y)] DPn CM ) ktrM/kp (transfer constant to monomer assumed to be that measured in CH2Cl2 (9 × 10-4)) Ctr ) ktr/kp (transfer constant to counterion) Y ) yield The corresponding graph is shown in Figure 8. The intercept, 3.1 × 10-3, is slightly higher than the concentration of cumyl chloride (2.5 × 10-3 M), and the value of the transfer constant to the counterion is 1.47 × 10-3. Because of the logarithmic term, this relationship can only be used for partial conversions. However, Figure 9 shows the calculated values for the evolution of the molar mass with conversion for a given value of the zero order transfer rate constant (Ctr ) 10-3) and of initiator (2.5 × 10-3 M) at variable initial monomer concentration. For any given yield lower than one, the molar mass of the polymer increases with the initial


Figure 8. Measurement of the zero order transfer constant in a compensated medium ( ) 5.4). [indene] ) 0.43 mol‚L-1; [SnCl4] ) 2 × 10-2 mol‚L-1; [CumCl] ) 2.5 × 10-3 mol‚L-1. Temperature: -15 °C.

Figure 9. Calculated values for the evolution of the molar mass with conversion at variable initial monomer concentration in the case of a zero order transfer. [M]o ) 1 mol‚L-1; [I]0 ) 2.5 × 10-3 mol‚L-1; ktr/kp ) 10-3 s-1.

monomer concentration. This is in agreement with the results of Table 3 ( ) 7.2; variable monomer concentration, complete conversion). However, this does not account for the decrease of the molar masses observed in the case of the polymerization of indene in pure CH2Cl2. A possible explanation for this decrease may be the time span of the polymerization. In a medium of low polarity ( ) 5.4) the reaction is slow enough, the heat exchange can take place, and the polymerization is isothermal, as we have verified. In CH2Cl2, the reaction is very fast and the temperature increase due to the heat of polymerization may favor transfer reactions, if heat exchange between the reaction vessel and the thermostated bath is slow. The heat release (estimated at about 50 °C for adiabatic conditions for [indene]0 ) 2.5 mol‚L-1) would enhance transfer reactions, and this effect would add up to the “normal” transfer and decrease the molar masses. Furthermore, the hypothesis of a zero order transfer is in agreement with the experiment at high indene concentration (2.5 mol‚L-1) in the presence of an hindered base. The polymerization stops at 11% yield. The Mn (12 000) corresponds to that of a polymerization at 0.25 mol‚L-1 initial indene concentration carried out to complete conversion, i.e., in a range in which the Mn is


controlled. Transfer to the monomer, even if it takes place to some extent, should not lead to a termination by the base.26,27 Consequently, the termination observed here should result from a zero order transfer with subsequent proton capture by the base. In the absence of base, the reaction goes to completion to yield a polyindene with the same molar mass. In a medium of dielectric constant 5.4, bimodal MWDs are observed, both at low conversion and at low initial monomer concentration, and the two peaks merge afterward. In the case of polymerizations in pure CH2Cl2 (or for ) 7.2) the same phenomenon might take place but would not be detected because conversion is complete. This bimodality has been attributed to propagations involving two different active centers, with little interconversion during the polymerization. The two possibilities are either separate propagation on free ions and ion pairs or initiation by adventitious protic initiator present in the medium (presumably HCl, which could come from SnCl4 or residual H2O). HCl is an efficient initiator with SnCl4.20 However, initiation by an adventitious HCl would in principle yield the same counterion (assumedly SnCl5-) than initiation by cumyl chloride, and there should be no difference between the behavior of the two types of initiation. A faster initiation by HCl than with CumCl, which would yield high molar mass polyindene at the beginning of the polymerization, can be ruled out because the ratio of the peaks of high and low molar masses remains constant during the reaction and the two polymers are formed simultaneously. On the other hand, H2O is not an efficient initiator with SnCl4,20 and it should be present in very low concentration given the experimental conditions. In the present case, given the low dielectric constant of the medium ( ) 5.4), the contribution of free ions might at first sight appear less likely, unless the equilibrium constant between active and dormant species is low enough to yield a small concentration of ion pairs, leading to important dissociation. On the other hand, theoretical calculations concerning the evolution of the MWD with yield for polymerizations involving ion pairs and free ions yielded bimolar distributions for the polymerization of styrene at -20 °C initiated with phenylethyl chloride and with SnCl4 as activator, in fairly good agreement with experimental results.28 In the present case, the Mn of the high polymer decreases with yield and that of the low polymer increases (Table 4), which is in agreement with this theoretical model. As initiation with HCl (which could be present in the SnCl4) should not lead to bimodality and H2O (which should be present in very low concentration given the experimental conditions) is not an efficient initiator with SnCl4, the most likely explanation is a propagation on both free ions and ion pairs. Assuming that the apparent rate constants for the two species have similar values, the bimodality may result from the different lifetimes of the species, the free ions being destroyed through a bimolecular reaction, and the ion pairs, through a unimolecular one. In the compensated medium ( ) 5.4; [indene]0 ) 0.43 M), addition of 2,6-di-tert-butyl-4-methylpyridine suppresses the high Mn peak. This may be due to a common ion effect of pyridinium pentachlorostannate resulting from reaction of SnCl4 with a protic impurity which would shift the equilibrium in favor of the ions pairs.

716

Givehchi et al.

Acknowledgment. We want to thank Professor Krzystof Matyjaszewski for helpful comments and suggestions. References and Notes (1) Miyamoto, M.; Sawamoto, M.; Higashimura, T. Macromolecules 1984, 17, 265. (2) Faust, R.; Kennedy, J. P. Polym. Bull. 1986, 15, 317. (3) Matyjaszewski, K.; Sigwalt, P. Polym. Int. 1994, 35, 1. (4) Cationic polymerizations; Matyjaszewski, K., Ed.; Marcel Dekker: New York, 1996; p 265. (5) Cationic polymerization, Fundamentals and applications; Faust, R., Shaffer, T. D., Eds.; ACS Symposium Series; American Chemical Society: Washington, DC, 1997. (6) Matyjaszewski, K.; Lin, C.-H.; Pugh, C. Macromolecules 1993, 26, 2649. (7) Matyjaszewski, K. J. Phys. Org. Chem. 1995, 8, 197. (8) Sigwalt, P. Makromol. Chem., Macromol. Symp. 1991, 47, 179. (9) Sigwalt, P.; Polton, A.; Tardi, M. J. Macromol. Sci., Pure Appl. Chem. 1994, A 31, 953. (10) Kennedy, J. P.; Midha, S.; Keszler, B. Macromolecules 1994, 26, 424. (11) Thomas, L.; Tardi, M.; Polton, A.; Sigwalt, P. Macromolecules 1992, 26, 4075. (12) Thomas, L.; Polton, A.; Tardi, M.; Sigwalt, P. Macromolecules 1993, 25, 5886. (13) Thomas, L.; Polton, A.; Tardi, M.; Sigwalt, P. Macromolecules 1995, 28, 2105.

Macromolecules, Vol. 33, No. 3, 2000 (14) Ishihama, Y.; Sawamoto, M.; Higashimura, T. Polym. Bull. 1990, 24, 201. (15) Lin, C.-H.; Xiang, J. S.; Matyjaszewski, K. Macromolecules 1993, 26, 2785. (16) Ishihama, Y.; Sawamoto, M.; Higashimura, T. Polym. Bull. 1990, 23, 381. (17) Kwon, O.-S.; Kim, Y. B.; Kwon, S.-K.; Choi, B.-S.; Choi, S.K. Makromol. Chem. 1993, 194, 251. (18) Kwon, O.-S.; Gho, G.; G.; Kwon, S.-K.; Choi, B.-S.; Choi, S.K. Macromol. Chem. Phys. 1994, 195, 2187. (19) Yang, M. L.; Li, K.; Stover, D. H. Macromol Rapid Commun. 1994, 15, 425. (20) Polton, A.; Sigwalt, P. Bull. Soc. Chim. Fr. 1970, 131. (21) Sigwalt, P. J. Polym. Sci. 1961, 52, 15. (22) Mare´chal, E.; Evrard, P.; Sigwalt, P. Bull. Soc. Chim. Fr. 1969, 1981. (23) Matyjaszewski, K.; Lin, C. H.; Bon, A.; Xiang, J. S. Macromol. Symp. 1994, 85, 79. (24) Tazi, M.; Tardi, M.; Polton, A.; Sigwalt, P. Br. Polym. J. 1987, 19, 369. (25) Leleu, J.-M.; Tardi, M.; Polton, A.; Sigwalt, P. Macromol. Symp. 1991,47, 253. (26) Fodor, Z.; Bae, Y. C.; Faust, R. Macromolecules 1998, 31, 4439. (27) Guhaniyogi, S. C.; Kennedy, J. P.; Kelen, T. J. Macromol. Sci.sChem. 1982, A(18), 77. (28) Matyjaszewski, K.; Szymanski, R.; Teodorescu, M. Macromolecules 1994, 27, 7565.

MA9913010

Influence of Monomer Concentration and Dielectric Constant on

Recommend Documents